Brain Tumors
Primary tumors of the brain can arise from different types of cells in the central nervous system. Medulloblastomas are derived from precursors of neuronal cells while astrocytomas are derived from the astrocytic subset of glial cells, and oligodendrogliomas are derived from the oligodendroglia precursor subset of glial cells. Other types of primary tumors are derived from cells that form the inner and outer linings of the brain such as ependymomas from ependymal cells, and meningiomas from cells that comprise the meninges, respectively. Glioblastoma multiforme (GBM) derived from astrocytes is the most common and deadliest primary brain tumor and is therefore classified as astrocytoma WHO Grade IV. (1)
The current standard of care for GBM is an aggressive surgical resection followed by radiation treatments and chemotherapy. Even with advancements in intra-operative brain imaging, which have assisted with gross resections, new chemotherapies, and more focused radiation treatments, the prognosis for GBM remains extremely poor with survival rates of 33% for one year and a five year survival rate of 5%.(2) Advances in treating GBM have made great progress in recent years. Much of this progress is due to a greater understanding of the molecular subtypes of GBMs, the elucidation of glioma stem cells (GSC) and their role in self-renewal and resistance to therapy (3-5), and advances in biological therapies.
Molecular Classification of Glioblastoma Multiforme
In 2010 Verhaak et al. classified GBM into subtypes based on clinically relevant characteristics. These subtypes are classical, mesenchymal, proneural, and neural. (6) This information has furthered The Cancer Genome Atlas (TCGA) and may provide greater insight in future GBM treatment. Classical GBM is characterized by an increase amplification of epidermal growth factor receptor (EGFR), frequently has amplification of chromosome 7 and loss of chromosome 10. (6)
Patients with classical tumors benefit the most from a combination of temozolomide (TMZ) chemotherapy with radiotherapy. (7) Furthermore, the retinoblastoma protein (Rb) pathways are altered by CDKN2A deletion. (6) The mesenchymal group typically coincides with a decrease in neurofibromatosis 1 (NF1) expression and co-mutations of NF1, PTEN and AKT. (6) There is also an increase in necrosis and inflammation. (6) The proneural group is distinguished by alterations in PDGFRA and point mutations of IDH1.(6) Unlike the classical subgroup, the proneural subgroup has less prevalent amplification of chromosome 7 and loss of chromosome 10. (6) The proneural subgroup also displays the most CD133 expression, a GSC signature. Even though the proneural subgroup shows a high CD133 signature, the overall survival of these patients has been significantly higher. (7, 8) Finally, the neural subgroup exhibits expression of NF1, GABRA1, SYT1 and SLC12A5. (6) Verhaak et al also discuss how aggressive treatment is shown to be beneficial for classical and mesenchymal subgroups, effective for neural subgroup, and no change for the proneural subgroup. (6) One thought is the proneural subgroup does not respond to more aggressive treatment because of the amplification of CD133 which possesses cancer stem cell properties.(3)
The methylation of the MGMT promoter is another key prognostic factor in GBM. MGMT hyper methylation has been shown to be associated with increases in long term survival by 77.8%. (9) Additional studies by Noushmerher et al. identified glioma-CPG island methylation phenotype (G-CIMP) in glioblastoma and low-grade gliomas based on concentrated hypermethylation. Patients with G-CIMP tumors displayed IDH1 somatic mutations classifying them within the proneural subgroup. Furthermore, patients with G-CIMP were typically younger at the time of diagnosis and had an improved outcome which is consistent with the proneural subgroup classification. (6, 10)
An alternative method to classify brain tumors is according to the patient's age, tumor volume, and Karnofsky performance status (KPS). Zinn et al. evaluated patient outcomes according to those four categories. (8) Patients were given one point for presenting each of the following characteristics: over the age of 60, KPS of less than 100, or tumor volume greater than 30,000 mm3. (8) Patients were then grouped according to their score, VAK-A for 0-1 points, VAK-B for 2-3 points. They concluded that VAK-A patients had a more favorable outcome considering they typically presented with greater p53 activation, longer median overall survival (20 months versus 12 months,) and displayed MGMT promoter methylation. (8)
Our ability to stratify GBM patients based on molecular biomarkers allows the identification of long-term vs. short-term survivors after treatment with current standard of care. Nevertheless, a 5% “long-term” survival rate for GBM patients 5 years after diagnosis is a poor prognosis. Biological therapies have now emerged, however, to provide possible curative approaches for treating malignant brain tumors.
Immunotherapy
Cancer immunotherapy is the stimulation of one's own immune system to activate specific immune cells to target and attack cancer cells. A major benefit of immunotherapy is its activation of immune surveillance against metastatic tumor behavior. (11) Additionally, it elicits adoptive immunity against cancer cells and memory T cells for recognition of recurrent tumor cells that exhibit previously expressed tumor antigens. The central nervous system (CNS), specifically the brain, was once considered immunoprivileged, however, studies have shown the brain is not a completely privileged site. Under normal conditions activated T-cells have been shown capable to cross the blood brain barrier (BBB) into the parenchyma. Nevertheless, studies also suggest that T-cells are required to be activated before they are capable of penetration, which is supported by studies showing that non-activated T-cells are limited on their ability to cross into the CNS. (12-14)
Furthermore, macrophages appear to be in abundance in CSF. (12) Radioactive antibodies have also been used to show a greater uptake of antibodies in conditions of brain cancers. This suggests that during a neuro-inflammatory disease state (e.g. encephalitis, meningitis, and cancer) antibodies and immune cells are more permeable to the BBB. (11) This gives rise to the potential use of immunotherapy against glioblastoma multiforme and other high grade gliomas. Immunotherapy methods currently used against high grade gliomas include, but not exclusive to the use of: activation and transfer of dendritic cells, adoptive T cell transfer, cytokines, and viral vaccines. (15)
Dendritic Cell Therapy
Dendritic cells (DCs) are so-called professional antigen presenting cells and are highly specialized in antigen processing, presentation, and playing a crucial role in adoptive immunity and immunological memory. (16) Even though DCs only represent a small fraction of circulating leukocytes, they play a major role in the immune surveillance. (17) DCs process and present antigens through the MHC I and II cell surface molecules and can activate both CD4+ and CD8+ T-cells. (16, 17) They can also assist in stimulating B-cells, and follicular DCs have also been shown to play a role in maintaining B cells memory. (16) DC vaccines use DCs loaded with antigens with the goal of initiating a T-cell antitumor response. (17)
After early animal models using DCs vaccinations demonstrated to be efficacious, safe, and capable of protecting mice re-challenged with tumors (18-20), the first pilot study using DC vaccination against brain cancer was performed at UCLA in 1997. (19-22) The first patient's time to progression following DC vaccination was 2 months with an overall survival of 21 months, displaying no significant adverse effects from the vaccine. (21) As a continuation of this pilot, Liau et al studied autologous tumor lysate (ATL) pulsed DCs in an intracranial rat model and demonstrated prolonged survival in the treatment group compared to the control, leading to a phase I trial; this phase I trial demonstrated efficacy in humans. (19, 23, 24)
In a more recent phase I clinical trial, Prins et al. examined the use of ATL-pulsed DCs versus glioma-associated antigen (GAA) pulsed DCs. (25) They concluded that the ATL treatment group had an increased median survival of 34.4 months compared to 14.5 in GAA. However, due to the high variance in the patient populations a meaningful comparison could not be made. (25) This study provides a look into the increased benefit of patient-specific therapy. Since then, many more DC clinical trials have been performed as shown below in Table 2, and there are approximately seven clinical trials currently active on Clinical Trails.gov. DC vaccines are shown to be safe with minimal adverse effects, with a benefit to overall survival, and allow for patient specific therapy making DC vaccination a compelling immunotherapy.
Vaccination with Tumor Specific Antigens
Epidermal Growth Factor Receptor variant III (EGFRvIII) is a tumor specific mutation commonly found in malignant gliomas. (26) The EGFR mutation encodes an active tyrosine kinase which leads to an increase tumorgenicity and migration, and patients exhibiting this mutation commonly results in poor prognosis. (26) This mutation however forms a tumor specific epitope which allows for immunotherapeutic targeting.(26, 27) In one of the first preclinical animal studies published using an EGFRvIII peptide conjugated with keyhole limpet hemocyanin, 70% of the treatment group displayed impalpable subcutaneous tumors, and in the intracranial tumor model the median survival increased by greater than 173% with 80% long term survivors. (28) This in vivo study showed the immune response was directed by NK and CD8+ T cells. (28) These early animal studies validated EGFRvIII vaccinations as a promising therapeutic and lead to multiple clinical studies.
Four clinical trials have been performed by using rindopepimut (CDX-110), an EGFRvIII specific peptide vaccine. (29) These clinical trials have shown both efficacy and the safety of this vaccine, leading to a current phase III trial in newly diagnosed GBM and a phase II in recurrent GBM. (29) An important EGFRvIII study was a multicenter phase II clinical trial using the EGFRvIII vaccine concurrently with TMZ, although TMZ induces lymphopenia, the EGFRvIII vaccine was able to induce an immune response eliminating EGFRvIII expressing tumor cells. (30) However, the phase III study, Act IV, was discontinued after interim analysis revealed no statistical benefit in overall survival. (31)
One difficulty with using EGFR inhibitors has been the development of tumor resistance. Sampson et al. found that of the twelve recurrent tumors analyzed after rindopepimut treatment, eleven had lost expression of EGFRvIII. (30) They suspect the loss of EGFRvIII expressing cells is due to the targeted therapy and clearance of these cells by the immune system. (30) However, Sequist et al. analyzed biopsies of 37 non-small cell lung cancer patients that had developed drug resistance to EGFR treatment.(32) Out of the 37 patients, 49% developed resistance through a T790 mutation, 14% from SCLC transformation, 5% from a MET amplification, 5% from a PIK3CA mutation, and finally 30% from an unknown mechanism. Understanding these tumor resistance mechanisms may give rise to using a combination of agents and better post resistance therapeutic treatments such as Engelman's combination of a MET inhibitor and EGF inhibitor.
Adoptive T Cell Transfer
Adoptive cell therapies are the largest division of immunotherapy for GBM and are continuing to grow. (33) Adoptive cell transfer (ACT) is a treatment which uses anti-tumor T-lymphocytes, autologous or allogeneic, expanded in vivo, then reinfused post lymphodepletion. (34) Currently the use of autologous tumor infiltrating lymphocytes is the most effective treatment for metastatic melanoma, resulting in 50% objectively responding and some patients exhibiting complete responses. (34) In an early clinical trial performed by Katakura et al 5 newly diagnosed GBM patients underwent ACT using γδT lymphonkine—activated killer (LAK) cells cultured in anti-CD3 mAb-coated flasks containing recombinant interlukin 2 (IL2)(35) This trial resulted in a median progression—free survival of 88 months, a median overall survival of 96.8 months, and no toxicity greater or equal to grade 2.(35) While this was a small clinical trial, this appears to be a promising approach in immunotherapy against GBM.
ACT has the capability of engineering T-lymphocytes to recognize a variety of antigens, making it useful for a variety of cancers. (34) The idea of engineering T-lymphocytes is a rapidly emerging area of research. One approach in which T-cells can be engineered is in the way they recognize antigens.
Two forms of antigen recognition being explored for ACT are T-cell receptors (TCRs) or chimeric antigen receptors (CARs). One of the major differences between CARs and TCRs is how they recognize their targets. (36) TCRs recognize antigens through major histocompatibility complex (MHC). Two restrictions of TCRs are autoimmunity from cross reactivity and the limitation by MHC complex formation on the number of potential targets.(37) On the other hand, CAR therapy is MHC independent and directly recognizes the target antigen.(36) CAR therapy is directed through antibody recognition and can be targeted towards any cell surface antigens. Some of the advantages of CAR therapy over TCRs are the number of antigens capable of targeting and high affinity targeting. (37) The higher affinity targeting of CAR therapy allows for less unforeseen autoimmune toxicity compared to TCRs. (37)
The outlook of CAR T-cells appears increasingly promising after Singh and Cooper (38) combined the novel Sleeping Beauty (SB) transposon/transponase system, developed by Hackett and colleagues (39), with T-cell engineering. Using SB to engineer CAR T-cells not only avoids using viral DNA, but has demonstrated to be faster and far less expensive compared to its viral recombination counterpart. (38, 40) In an ongoing phase I/II clinical trials, SB CD19 CAR T-cells have shown to be effective, resulting in all 4 patients with Non-Hodgkin's lymphoma remaining in remission after 3 months post infusion, and 1 acute lymphoblastic leukemia patient remaining in remission after 5 months. Additionally, there have been no acute or late toxicities to date. (41) Currently, there are three clinical trials activity recruiting patients to investigate SB CAR engineered T-cell usefulness in lymphoma and leukemia. In the context of brain cancer, there have been several preclinical studies using chimeric antigen receptors that have shown efficacy and currently an ongoing pilot study for autologous EGFRvIII CART cells (NCT 02209376).(42-51)
A second approach in which T-cells can be engineered are in the antigens they are specified to recognize. While T-cell engineering opens the door for targeting a large variety of antigens, emphasis should be place on choosing antigens not expressed by healthy tissue as autoimmune toxicity still remains as a major limitation to ACT. Likewise, the variability in response to ACT by patients makes it that much more difficult to predict unanticipated autoimmune adverse events.(36)
According to Hinrichs the four types of antigens that are ideal for targeting by ACT are those that are: expressed predominantly by germ cells, mutant gene products not expressed by healthy tissue, antigens restricted to nonvital tissue or cell lineages, or finally viral antigens.(36) Under these guidelines, some antigens that emerge as targets for gliomas include EGFRvIII, IL13Rα2, and human cytomegalovirus (HCMV). EGFRvIII is a suitable ACT target since it is a mutant gene not expressed by healthy tissue, and produces a novel peptide that is not normally found in the body. There have been several preclinical studies have been performed indicating the effectiveness of EGFRvIII CAR transduced T-cells.(47-49, 52) Miao et al. have displayed EGFRvIII CAR T-cell migration across the BBB and localization to the tumor region. (47) Moreover, Sampson and colleagues demonstrated that EGFRvIII CAR modified T-cells cured tumor bearing mice and generated anti-tumor immunity as evidenced by resisting tumor rechallenge with EGFRvIIINeg tumors. (49) These pre-clinical studies have demonstrated the feasibility and safety of EGFRvIII CAR mediated T-cells for clinical trials, which have already begun enrolling.(48)
Another glioma specific target for CAR T-cells is IL13Rα2. IL13Rα2 is a highly overexpressed cell surface receptor expressed in approximately 58% of adult gliomas. (46, 53) Thus far two clinical trials have been performed using IL13-zetakine CAR directed T-cells, which demonstrated feasibility without serious toxicity.(53) The overall survival and time to progression have not yet been published. However, one of the participants underwent a whole body and brain PET scan to detect localization of infused T-cells to the tumor regions.(54) It is anticipated that a second generation IL13Rα2 CAR T-cell clinical trial that will begin recruiting in the near future.
Human cytomegalovirus protein has been found in 90-100% of primary GBM making it an appealing target for ACT, but thus far only a handful of CMV-specific ACT clinical trials have been performed. (55) In a clinical trial using CMV-targeted T-cells, 10 of the patients recruited received a minimum of three infusions. Of these 10 patients 4 were progression free through the duration of the study. (56) The median overall survival was 403 days from the first recurrence, with minimal toxicity reported.(56) Additional clinical trials are needed to confirm efficacy and safety of CMV-specific ACT. One difficulty that has been noted for CMV immunotherapies has been overcoming the immunosuppressive tumor microenvironment.(55)
An alternative to targeting glioma specific antigens is targeting non-specific tumor associated antigens such as those prioritized by the National Cancer Institute for therapeutic targets and cancer vaccines. These antigens include: WT1, HER2, p53, MART1, gp100, the MAGEs and several more. (57)
The National Cancer Institute prioritization project ranked WT1 first amongst the other antigens used for therapeutic targets and cancer vaccines. (57) Wilms' tumor protein 1 (WT1) was previously described as a tumor suppressor gene, but is now shown to act as an oncogene. (58, 59) WT1 is a promising target because it is highly expressed in several solid tumors and hematological malignancies while only being expressed in a few normal tissues. Importantly, Driessche et al reviewed all reported WT-1 targeted immunotherapy cancer vaccine trials and found objective clinical responses in 46% of all solid tumors, and 64% in all hematological malignancies.(58) WT1 remains to be a safe therapeutic target, with only two patients having had severe adverse events, both patients had myelodysplasia syndrome.(58)
In contrast, MAGE-A3, a cancer testis gene, showed major cytotoxicity in a TCR-ACT clinical trial. Three of the nine patients treated with autologous anti-MAGE-A3 TCR engineered T-cells exhibited changes in mental status, two of which lapsed into comas and died. (60) Researchers suggest the initiating event was unrecognized MAGE-A12 expression in the brain. (60) Another common target of immunotherapies is human epidermal growth factor receptor 2 (HER2/ERBB2). HER2 is a growth receptor amplified in several cancers, including GBM, but is most commonly known for its amplification in breast cancer.(61) HER-2 has yet to be tested in a clinical trial against GBM, but several preclinical studies have been performed in medulloblastoma and GBM.(51, 62, 63) There have been no serious autoimmunity adverse events shown in the vaccine trials performed targeting HER2, yet cytotoxicity remains a concern due to the fact it is commonly expressed in vital tissue such as the heart, lungs, kidney, and bowels.(36) One method being explored to limit cytotoxicity from antigens expressed in healthy tissue is by engineering bispecific CAR T-cells. (64) In addition to limiting toxicity, bispecific CAR T-cells may provide a way to overcome recurrence from antigen null tumor cells that escape antigen targeting.(64) Using a binomial mathematical model, Hegde found dual targeting to be far superior to single antigen targeting in achieving near complete tumor cell capture. (64)
Gene Therapy
Gene therapy has been tested in numerous phase I clinical trials against brain tumors as well as pancreatic, lung, prostate, and renal carcinomas. While some forms of gene therapy have shown to be effective in certain cancers such as malignant melanoma and pancreatic cancer, few GBM clinical trials have advanced past phase II. (65) Like all other immunotherapy treatments, the goal of gene therapy is to provoke an immune response to destroy cancer cells. In a preclinical trial by Kuriyama et al HSV-tk-transduced murine hepatocellular carcinoma (HCC) cells were implanted subcutaneously into immunocompetent BALB/c mice and BALB/c nude mice. (66) Both groups were then treated with ganciclovir, however, only the immunocompetent mice displayed inhibition of tumor formation.(66)
Furthermore, the immunocompetent mice exhibited infiltration by lymphocytes including CD4+ and CD8+. (66) This demonstrated that T-cell-mediated immune responses play critical role in HSV-tk gene therapy against cancer.(66) Many of the GBM trials have used either the adenoviral vector containing the herpes simplex virus thymidine kinase gene (AdV-tk), or herpes simplex thymidine kinase gene (HSV-tk) vectors as shown below in Table 3. These vectors have been used to deliver the p53 gene, a cell regulator and commonly causes apoptosis to cancer cells.(65) The combination of TK and granciclovir creates cytotoxic nucleotides selective for dividing cells and stops DNA synthesis.(67) The first measurable survival improvement trial was performed by Immonen et al using a combination of AdvHSV-tk and ganciclovir which resulted in median survival of 62.4 weeks in the treated group compared to 30.9 weeks in the control.(65, 67) While this study showed to be effective, many other clinical trials including a large phase III have not shown to be as effective. One of the problems with the HSV-tk retroviruses has been low penetration of brain tissue and low transduction efficiency.(67)
An alternative approach to inducing suicide gene therapy is using mesenchymal stem or neural stem cells expressing HSV-tk. Both mesenchymal stem and neural stem cell vehicles have been shown in preclinical studies to be safe and effective in achieving the bystander effect of suicide gene therapy. (68-72)
Cytokine gene vaccination is yet another form of gene therapy. Cytokine vaccinations consist of using cytokines or recombinant cytokines to elicit an anti-tumor response by the immune system. Unfortunately, many cytokines induce toxic reactions and are unstable in vivo. However, many delivery vehicles have been tested to overcome these adversities some of which include: adenoviral delivery, modified neural stem cells, or genetically engineered tumor cells. (19, 22, 73-75) The more commonly tested cytokines exhibiting efficacy against malignant gliomas are: granulocyte macrophage colony stimulating factor (GM-CSF,) Interferon alpha, and Interlukins IL-2, IL-4, and IL-12. Additionally, cytokine therapy can be used in combination with other immunotherapies such as DC, tumor lysate or tumor antigen vaccinations.(76-79) In one pre-clinical study GM-CSF with irradiated tumor cells increased the survival rate of intracranial tumor bearing rats compared to the control rats. (80) This study also showed combining IL-2 or IL-12 to the therapy lead to in an increased survival rate up to 90%.(80)
Oncolytic Viruses
Rather than using replication-incompetent viral vectors to deliver genes as discussed above with gene therapy, oncolytic virus therapy employ viruses with an active life cycle. Although both methods introduce viruses to kill cancer cells without harm to healthy cells, oncolytic viruses also use the patient's immune system to further attack the cancer cells enhancing the effectiveness of the treatment. (81) These tumor-selective viral replications results in lytic tumor cell destruction and subsequent release of thousands of vial progeny that go on to infect neighboring tumor cells. As a result, this method practices a local self-amplification therapeutic effect that is unique in comparison to all other forms of treatment.
There is significant concern in using replication competent lytic viruses in the brain which is heightened when considering the injection of pathogenic viruses. However, oncolytic viral therapy is particularly attractive to GBM patients because the tumor is confined to one organ and the tumor cells grow surrounded mostly by post-mitotic cells. Therefore, there is a reduced risk of the treatment damaging normal surrounding cells. Even so, only highly attenuated agents, or viruses that have substantial genetic modification deleting viral genes of harm, are typically considered. (81) Selective viral attenuation is typically considered to be vital to oncolytic viral therapy to create not only safer viruses but also increased tumor selectivity. A number of viruses have been modified and evaluated for their oncolytic potential. These include polio virus, herpes simplex virus, adenovirus, and measles virus.
Poliovirus
One of the most promising oncolytic viruses is the poliovirus. It has shown strong potential in its ability to target tumor cell killing and engage the host immune system. The potential of poliovirus stems from its unique mechanism of invading the host organism. The positive stranded RNA virus is a natural neuropathogen with neuroinvasiveness in particularly making it an ideal candidate for effective treatment against GBM. Early in the viral life-cycle, viral 2A protease is expressed and engages in rapid cleavage of key host cell components involved in mRNA export and translation. By disengaging the eukaryotic initiation factor (eIF)4G and the nuclear pore complex, within 2-3 h the virus has shut down the host cell gene expression limiting antiviral responses that require biosynthesis. (82)
What is even more significant is that this does not affect the virus gene translation process. Simultaneously, in the absence of intact eIF4G viral replication and translation continues. Whereas usually, eIF4G is needed along with eIF4E to cap the 5′end of mRNA before undergoing translation, poliovirus uses a cis-acting genetic element in its 5′ IRES to recruit ribosomal subunits. Therefore, its mechanism of attack overrides the host cell without engaging in complex parasitic relationships as most viruses do, allowing the virus to survive, replicate, and kill.
Its neuroinvasive quality stems from its ability to bind to the poliovirus receptor (PVR), commonly known as Necl-5 or CD155. These receptors are expressed on motor neurons. The binding of the two triggers receptor mediated endocytosis followed by a conformational change in the viral particle. (83) With this conformational change, a hydrophobic region becomes exposed on one of the capsid proteins enabling insertion into the endosomal membrane to form a pore where the viral genome can enter. Functional studies have since implicated Necl-5 involvement in cell invasion and intracerebral dispersion in glioblastomas. Furthermore, immunohisochemical studies have located the molecule in tumor cells at the invasive front of tumors. (84) Fluorescence-activated cell sorting and immunohistochemical studies of GBM patients' tumors confirmed universal and abundant expression of Necl-5. (85) Therefore, using poliovirus in oncolytic viral therapy may be the key to treating GBM.
In order to curtail the pathogenicity factor of the virus, the poliovirus IRES was genetically modified to be exchanged with a non-pathogenic version from human rhinovirus type 2 (HRV2). The chimera, RIPO, was then further modified to maximize attenuation. PVSRIPO was designed containing live attenuated SABIN poliovirus vaccine. Each vaccine contains a single point mutation in their respective IRES elements located in a distinctive stem loop domain, V, to prevent poliomyelitis. Overall, tumor specificity is then based on the affinity for the PVR that is upregulated in neruoectodermal malignancies and on functional growth deficit of the HRV2 IRES element in normal cells of neuronal derivation. (82) Together, with the attenuation SABIN vaccine strain, tumor specificity is achieved while highly reducing toxicity potential.
In previous work, exchanging the complete poliovirus IRES for the human rhinovirus type 2 (HRV2) IRES generated the chimera PV-RIPO that was shown to depresses viral translation and propagation in neuron-like cell lines (e.g., Sk-N-Mc and HEK-293).(86, 87) In addition, while the heterologous HRV2 IRES prevents virus propagation in spinal cord motor neurons without causing poliomyelitis, it has shown no effect on rapid viral growth in non-neuronal malignant cell types like those derived from malignant glioma cells.(88-90)
The genetic stability of the oncolytic non-pathogenic poliovirus recombinant was examined in vivo to be considered for therapy of recurrent glioblastoma multiforme. Bilateral HTB-15 xenografts were implanted in 12 athymic Balb/c mice to monitor tumor regression and enable virus recovery from the same animal. Ten days post inoculations, the median xenograft size had shrunk by 45%. (91) On day 20, the median xenograft size had shrunk by 45%.
Histopathology of xenografts from mice on day 20 (10 days post PVS-RIPO inoculation) showed identical responses in all 6 animals analyzed.(91) Advanced tumor cell lysis was observed with a bulk of the tumor no longer resembling the appearance of proliferating tumor. The majority of the tumor mass had vastly reduced cell content and was diffusely invaded by infiltrates. Because histopathology of xenografts 28 days post PVS-RIPO injection had essentially been replaced by a scar, it was assumed that the tissue rearrangement represented the host's reaction to tumor destruction leading to a transition towards scar formation. The histopathological analysis suggested that active PVS-RIPO replication in xenografts and a vigorous host response to the receding infected tumor induced complete tumor elimination resulting in scar formation.
Viral recovery has been analyzed in PVS-RIPO injected mice. Tumor lysates were first tested by plaque assay to confirm the presence of virus.(91) Xenografts from 6 animals 10 days post-inoculation were seen to contain infectious material while no infectious material was able to be recovered from tumors 28 days after inoculation of the virus. The latter was not surprising given the low amount of xenograft remaining at that interval, its histopathological appearance, and the very low vial titers in xenografts at 10 days post inoculation. Overall, these results suggest that PVS-RIPO is unable to persist in tumor once the supply of viable tumor cells has disappeared.
These in vivo xenograft experiments of tumor regression and virus recovery suggested a vigorous host response to viral tumor cell lysis evident by extensive infiltrative lesions within and surrounding the xenograft, perivascular cuffing, active tissue remodeling and, eventually, scarring. Furthermore, host inflammatory reaction contributes to virus removal once tumor obliteration eliminates the site of active viral replication. Together, PVS—RIPO is especially unique among oncolytic viruses for its safety and tumor-specific replication that partially relies on an abnormal environment for translation initiation in malignant glioma cells. These experiments support consideration of the agent to move to clinical trials against recurrent GBM.
Clinical trials of PVS-RIPO against recurrent glioblastoma brain tumors are ongoing at Duke University. A phase I trial (NCT01491893) was initiated to determine the maximally tolerated dose (MTD) or optimal dose of PVSRIPO when delivered intracerebrally by convection-enhanced delivery (CED) involving sterotactically placed catheters directly into the malignancy. Patients were infused with the virus over a span of 6.5 hours with a delivery rate of 0.5 ml/hr. Patients with Grade IV malignant glioma tumors were selected and subsequently analyzed post-injection for progression-free survival (PFS) and overall survival (OS). Results of the phase I study were reported at the 2014 American Society of Clinical Oncology meeting. At that time the phase 1 trial included 15 patients with recurrent supratentorial glioblastoma. The medial survival was 15.2 months and the 12-month survival rate was 70%. Furthermore, 18 and 24-month survival rates were 43.8% and 29.2%, respectively. In regards to safety, dose escalation through five levels eventually revealed dose-limiting toxicity at level 5. Adverse events considered potentially relevant to the study remained at or under grade 3. (92) Since this phase I report the FDA has granted PVSRIPO breakthrough status (93) thus accelerating its availability to other brain tumor centers across the country for treatment of patients with recurrent GBM.
Adenovirus
Adenovirus, a double-stranded non-enveloped DNA virus, is another oncolytic virus that has been widely studied for its potential in cancer treatment. In nature, the adenovirus is as a very common pathogen causing mild upper respiratory symptoms.(81) The adenovirus enters the cell by receptor mediated endocytosis. As the endosome matures and becomes more acidic, the virus goes through a process of un-coating steps to remove structural proteins from the capsid. (83) Some of these steps rely on a viral protease which becomes activated in the reducing environment of the endosome. One protein released from the capsid lyses the endosomal membrane, releasing the remainder of the virus into the cytosol. The virus then docks onto the nuclear pore complex and the DNA is transported into the nucleus to be transcribed.
In previous glioblastoma studies, the results for the traditional adenovirus serotype 5 (Ad5)-based vectors had disappointing results due to low expression of the adenovirus receptor (CAR) on GBM cells.(94) More recent modifications of the adenoviral vector system have been designed that combines the capsid of the wild-type Ad5 with fiber proteins of the adenovirus group serotype 35 (Ad35). This has shown to change the viral receptor from CAR to the human CD46 receptor which is up-regulated in tumor cells. (95) In addition, certain proteins from the E1A gene region in the adenovirus trigger cells to enter the S-phase in the cell cycle by interacting with cellular retinoblastoma tumor suppressor protein (pRb). Similarly, E1B proteins suppress apoptosis by binding and inactivating p53, thereby inhibiting its pro-apoptotic response. (96) Because of these interactions with vital tumor growth processes, these proteins are commonly targeted in treatment methodologies.
The adenovirus can be further modified to promote tumor-specific gene expression. By deleting all E1A and E1B genes within the adenovirus, the virus becomes replication incompetent in normal cells, cells without cell cycle dysfunction. This renders the adenovirus for oncolytic virus therapy tumor-selective. In one study, homologous recombination between inverted repeats (IR) in adenovirus genomes forms the basis of the design. Ad.IR vectors were modified to express tumor necrosis factor-related apoptosis-inducing ligand (TRAIL). (95) When the virus was evaluated, there was pronounced improvement in the effectiveness of targeting human glioblastoma cells through the CD46 receptor. Human glioblastoma cell lines U-87 MG, T98G, and SF767 all showed higher mean fluorescence with anti-CD46 antibody compared with an anti-CAR antibody, demonstrating the improved tumor-selectiveness of the adenovirus.
The oncolytic potential of Ad5/Ad35.IR-E1A/TRAIL has been evaluated in comparison to the traditionally studied adenoviruses (WTAd5, WTAd35, and Ad5/Ad35.IR-E1A/GFP). After virus infection, cells were observed for 6 days in vitro. For Ad5/Ad35.IR-E1A/TRAIL, all cell lines resulted in tumor cell death while little to no cell death was seen with the wildtype adenovirus. After injection of the wildtype adenovirus at the same dosage, cell death was still seen (less than 40%) but only in the SF67 cell line by WT-Ad35. U-87 and T98G cells showed no significant cell death after infection with WT-Ad5, Ad35, or Ad5/Ad35.IR-E1A/GFP. These results suggest the newly developed adenovirus system was more effective at inducing cell death attributing the finding to the apoptosis-inducing gene TRAIL key role.
Wohlfahrt et al (2007) also analyzed the comparison between Ad5.Ad35.IR-E1A/TRAIL and AD5.1R-E1A/TRAIL to assess the antitumor effect with the combination of Ad5 and Ad35 as a result of the receptor switch from CAR to CD46. Cells were infected with either Ad5.Ad35.IR-E1A/TRAIL or AD5.1R-E1A/TRAIL and apoptosis was observed using TUNEL assay. By the end of day 4, 20% to 30% more cells infected with Ad5.Ad35.IR-E1A/TRAIL underwent apoptosis for all cell lines. In contrast, no elevated levels of apoptosis were detected in wild-type Ad35 or Ad.5.IR-E1A/TRAIL. These results suggest that Ad5 combined with Ad35 was more effective at inducing apoptosis most likely due to altering the primary receptor of the virus.
Tumor growth was compared to the size of tumors in vivo with U-87 MG-untreated mice. Beginning 8 days post-injection, growth impairment was seen in tumor cells infected with Ad5.Ad35.IR-E1A/TRAIL with 40% less tumor volume than the negative control. The average growth remained reduced with about 40% to 50% volume reduction compared to the untreated tumors for the entire 20-day follow-up. While slight growth impairment was also seen in the Ad5, Ad35, and Ad5/Ad35.IR-E1/GFP-treated tumors, on average tumor size was only reduced by 20% to 40%. Altogether, only tumors treated with Ad5/Ad35.IR-E1/TRAIL showed a significant effect on inhibiting tumor growth. Ad5 (also known as ONYX-15) was tested in a phase I clinical trial for glioblastoma and showed no significant toxic side effects.(97) With these results, investigators are advancing the Ad5/Ad35.IR-E1/TRAIL protocol due to its improved infection and enhanced apoptosis outcomes that could quite possibly prove to be a more effective oncolytic virus therapy treatment in clinical trials.(95)
Measles Virus
The measles virus is a negative stranded, enveloped, RNA virus that also demonstrates viral oncolytic lysis potential to treat brain tumors. There are two measles virus glycoproteins, the hemagglutinin protein H and fusion protein F, that are imperative for oncolytic specificity and efficacy.(98-100) Hemagglutinin protein H binds to its receptor to initiate fusion. Mutated viral H protein displays a high affinity to CD46 receptors which is overexpressed on numerous tumor cells this key to the successfulness of oncolytic measles viruses. Next, fusion protein F simultaneously activates processes leading to syncytia formation and ultimately apoptosis.
Preclinical studies have investigated the therapeutic effectiveness of a highly attenuated measles virus known as the Edmonston stain (MV-Ed) that was designed to express carcinoembryonic agent (CEA). (101) MV-CEA was first evaluated In vitro to evaluate its ability to replicate and induce tumor cell death. Because the CD46 receptor is expressed abundantly on these tumors, glioma cell lines U251, U87, and U118 were preferentially studied. In all three cell lines, cell death was observed in over 90% of cells by 72 hours after being infected with MV-CEA. Less than 1% of cells were viable by 120 h after infection regardless of the cell line. Cell death was achievable with MV-CEA.
MV-CEA replication in glioma cells was determined after titers of the virus from the cells were obtained 24, 48, and 72 hours post infection. An increase in titer was associated with a significant rise in the CEA level which corresponds with viral replication/gene expression. CEA levels in uninfected glioma cells were undetectable. The mechanism of cell death was analyzed through TUNEL assays in all three glioma cell lines. Early after infection (day 2), syncytia was TUNEL-negative. By day 4, the nuclear clusters in 50% of syncytia showed TUNEL-positive nuclear staining and the cytoplasm was invariably negative. After day 6, there was positive diffuse staining throughout the cytoplasm, with residual positive nuclei and nuclear fragments, thus confirming apoptosis as mechanism of cell death.
In vivo studies evaluated the antitumor effect of MV-CEA using subcutaneous U87 glioma xenografts in BALB/c nude mice. (101) All mice treated with MV-CEA, exhibited complete regression of subcutaneous tumors as well as a significant prolongation of survival. Mice injected with 8×107 pfu MV-CEA had significant tumor regression as compared to mice treated with the same dose of UV-inactivated MV-CEA or untreated mice. The antitumor effect of MC-CEA was also evaluated in a mouse model that more closely resembles the disease in humans with intracranial U87 xenografts in BALB/c nude mice. Significant regression of the intracranial tumors after administering 3×105 pfu/dose for 6 doses, totaling a dose of 1.8×106 pfu, of MV-CEA. For these intracranial tumors, 7 out of 8 mice that received the MV-CEA treatment had complete regression of the tumor based on Mitt A phase I study to test the safety of the measles virus for patients with GBM is currently underway at the Mayo clinic (NCT00390299).
Herpes Simplex Virus
As one of the first oncolytic viruses to be adopted to attack cancer cells, Herpes Simplex Virus (HSV), it one of the most well understood viruses, easiest to manipulate, and relatively harmless in nature posing fewer risks and making it an ideal candidate virus. These factors propelled the advancement of HSV into oncolytic virus therapy. Entry into the cell requires the coordination between the 4 enveloped glycoproteins gB, gD, gH, and gL. Viral infection is initiated as gB and gC bind to surface proteoglycans. Next, gD interacts with the specific surface receptor prompting a conformational change in the glycoprotein. This leads to fusion of membranes by the activation of gB and the gH/gL heterodimer. In GBM patients, several strains of the virus including G207, HSV1716, and NV1020 have had success in clinical trials demonstrating antitumor potential without toxicity. Each mutant features deleted or manipulated viral genes to reduce toxicity without interfering with the infection of actively dividing cells. (102) Although there have been some promising patient responses (NCT00028158, NCT02031965), the overall effectiveness of herpes simplex viral vectors for oncolysis of GBM has been limited.
Taken together, greater specificity of targeting and efficacy by oncolytic viruses is needed for the treatment of brain tumors.